Algorithm of Lord's Eye:

STEP - 1 Mount Google Drive

STEP - 2 Load the video Capture the frames (For our submission purpose, we have used 5) for slot detection. Find Frequency of cars and compare it with defined threshold to produce a map

STEP - 3 Display the rectangle drawn on detected slots.

STEP - 4 Save the co-ordinates in CSV file.

STEP - 5 Load the co-ordinates from CSV file Crop the given test image according to given RegionOfInterest

STEP - 6 Apply the model and get the results (center co-ordinates of the slot,occupied/not-occupied,color_code,pos) Show the test image with Red & Green rectangles with center points

NOTE: frozen_inference_graph.pb file being too large is not included in this repository (in mask-rcnn/mask-rcnn-coco folder)